The Application Of Data Warehouse And Data Mining In The Analysis Of College-wide Examination Course Grade

Data Warehouse and Data Mining technology has became the focus in the field of information application technology. They have been widely applied to finance, insurance, banking and so on, while there has been comparatively less application in education. For the goals to create a fully and improved knowledge systems for students, and prepare to cultivate the person of full development with talent, College Common Courses were opened in the university. At present, teachers usually adopt large classes with the College Common Courses, and students themselves do not very interest in the courses, so that it is difficult to improve the teaching level of the courses. College computer basic course is contained by the College Common compulsory Courses and the College-wide courses at the same time. In this paper, the technology of Data warehouse and Data mining were introduced to deeply analysis and research on the scores of college computer basic course, then scientific and effective data supports have been obtained to help college management department make their teaching plans.In this paper, through learning and research on the theory of data warehouse and data mining, combined with the investigation of our school computer basic testing system, and the analysis based on the scores of 7500 students which from 20 colleges in the two consecutive testing years, finally, we proposition that the technology of data warehouse and data mining have been applied in the analysis of computer basic wide scores. The method can applied in the other College-wide examination course, such as College English, Marxism philosophy and so on. This thesis falls into there mainly parts:Firstly, course grade data warehouse of college-wide examination have been build in the thesis. At first, data warehouse model have been build from three themes which contain analysis of total scores, difficulty of questions and scores of questions. Then, students'grades have been extracted from computer basic examination system, and have been transformed into pure data. Finally, pure data have been loaded in the mode of course grade data warehouse. On-line Analytical Processing technology which contains slice,dice,pivot and drill-up have been applied in this data warehouse to analysis students' grades from multi-layers and angles. The analysis contains total scores of students, difficulty of papers, question grades of students. Not only can the knowledge from the analysis help teaching management system to find out the trend of students'grades fluctuation, but also to explore rationality of difficulty distribution and to make out real cause of the tendency.Secondly, based on the course grade data warehouse of college-wide examination, association rule as a main technology of data mining have been used to analysis students' grades in depth level. In terms of computer foundation course, teachers pay different attention to grades of colleges, moreover, questions of chapters in test papers have different difficulty. Mixed weighted association rule have be used in the data warehouse to analysis of students' grades. During the process of analysis, the results which contrast with average and experiment vules from teachers have been considerated to determine the weight of item. Then a model and appropriate mining method of mixed weighted association rules for students'grades have been given in this paper. Compared to directly apply the Apriori algorithm, more valuable correlations have been obtained between the chapters, chapters and scores, colleges and chapters at the same threshold values. Not only can these knowledge as reference for teachers arrange the sequence of chapters content and reasonable teaching hours, but also guide teachers for putting the rational teaching plan to students from different colleges.Thirdly, decision tree have been applied to analysis students'grades of data warehouse. The research beginning with knowledge points of computer basic course. Because of the data were mass, sequence and so on, C4.5 algorithm have been adopt to build an effective classification model in this paper. The conclusions from analysis help us to discover a scientific approach of classification teaching, the approach beginning with the student knowledge points understanding. It can be used as a reference which can help improve teaching level of College Common Courses for teaching management system.Take the grade of college computer basic wide examination as a example, Through the mining practice in college-wide examination course grade, large valuable information has been extracted in this practice. These knowledge can assist teachers to make their teaching plan, enhance teaching level of college-wide examination course and guide teaching management department launch and improve education in the coming years.